Synthesis of "The OHOH" - a benzylic hydroxylation product

[david1597]

One of the potential identities of our mystery compound is where the benzylic position in the NW has been hydroxylated to give a tertiary alcohol. Concurrent to the continued investigations at Pfizer, I've set about making this compound to potentially give us some insight, one way or another. Aiming to complete within ten days to meet the World Malaria Day Dundee shipment to assess it's potency, but the NMR data itself could be super useful. This is definitely not the actual compound made in the biosynthesis - that was something para-substituted - but could still be very useful.

Proposed scheme below. Starts with a diol protection of Tris to give 1 followed by oxidative cleavage to the ketone 2. The triol 3 can then come about by a Grignard reaction, before the usual reaction onto the core to give "the OHOH", 4. Easy, right?

OHOH scheme.zip

[MFernflower]

Lovely work!!! @david1597 If you have free time might be neat to make the p-fluorophenyl version?
4-Fluorophenylmagnesium bromide is a common lab chemical is it not? That way we have something at the para location that we can pickout by hydrogen NMR?

Also I bet a neat cyclic ether type compound could be formed by reaction of 4 in a mix of crown ether, lye and 1,2 dichloroethane?

[mcoster]

Looks good - good luck @david1597

[MFernflower]

Just a harmless reminder @david1597 Can you add smiles and InChI data to your post?

[david1597]

So, it is now looking like this compound is the actual compound from that original biosynthesis. Things are changing quickly, hey? I wish I had written that last line of my opening paragraph a little differently now...

This is definitely not the actual compound made in the biosynthesis...

[drc007]

2,2-Dimethyl-1,3-dioxan-5-one is commercially available.
https://www.alfa.com/en/catalog/L20177/

This patent may also be of interest https://pubchem.ncbi.nlm.nih.gov/patent/US6255540

[MFernflower]

Dihydroxyacetone mixed with 2,2-Dimethoxypropane could work to form the precursor in a simple room temp (I think) synthesis with DCM and Zirconium(IV) chloride @david1597

http://www.organic-chemistry.org/abstracts/lit2/226.shtm

The dihydroxyacetone acetonide can be purchased but it's a tad bit pricey: http://www.matrixscientific.com/087907.html

[david1597]

Thanks @drc007 and @MFernflower. I went ahead with the original route as we had an abundance of those starting materials in the lab already, and it looked straightforward enough. Chemicals can take a while to order in here in Australia. I also wanted to make it fresh and use straight away as it's not that stable.

Good news is I have about 600 mg of 2,2-dimethyl-1,3-dioxan-5-one (plus 1 g+ more of 1) ready for the Grignard step today.

[MFernflower]

@david1597 Do you have 4 methoxy phenyl grignard and/or 4 fluoro phenyl grignard on hand?? Could be good derivatives to make?

[david1597]

I've now got the diol-protected triol following the Grignard reaction, 740 mg of it. The deprotection will be stirring overnight.

And yes, I'd love to make a couple more derivatives before the shipment in a weeks time. If the coupling of this one goes well, this should be possible. I'm making the Grignard's fresh, so anything we have an aryl bromide for is a possibility.

[david1597]

• the deprotection gave 550 mg of the triol, 3 - ELN

• half of this was taken into the final substitution with the core. Multiple products were obtained after purification. Mass spec indicates that the product may be in at least one of these. I'm currently drying these and will then obtain the NMR's - ELN.

• I'm still aiming to make another two derivatives. These being the 3,4-difluoro- [nine of our current top ten most potent hits contain this], and the 4-dimethylaniline [beginning to explore SAR with something electron donating]

[MFernflower]

RE: the 4-dimethylaniline
Do we know if compounds bearing this fragment are mutagenic in any way? aryldimethylamines can be monodemethylated so we aught to watch our footing on these molecules
https://www.ncbi.nlm.nih.gov/pubmed/2750202
@david1597

[david1597]

Unfortunately I did not get to the final compound in time to make the shipment to Dundee. Sorry team.

The final substitution onto the core proved tricky and the desired product has not been isolated. Other products have been - one may be an epoxide. It may be necessary to protect the triol, allowing more forcing conditions to improve the yield and improve isolation, followed by deprotection. I am now on leave for a few days and will resume the synthesis next week.

[MFernflower]

@david1597 Just a random Q - do you know what the epoxide might look like???

[david1597]

I'm currently getting NMR and mass spec data for the products that were isolated. Will get back with some answers within the next couple of days...

[david1597]

Discussion of the side-products formed during the final step of this synthesis will continue at #41 .

[david1597]

Given the difficulties in the direct coupling the triol I am attempting a route via acetonide protection. Using the N-dimethylaniline triol, the desired product where one of the secondary alcohols remains free was obtained (ELN link) in a 3:1 ratio over the free tertiary alcohol (which is the same compound as two steps previously...). Currently (ELN link) attempting to couple this to the core.

[MFernflower]

Very nice - The starting triol can be regenerated from the waste product making this a truly green synthesis!

[david1597]

Quick update before the weekend:

• using the protected alcohol has worked (by 1H NMR and MS) - DGS-69-1. However, only a small amount of material obtained so far. I was observing three new spots by tlc, and cautious about getting garbage again, stopped the reaction to see what they were. They were the -OH and -OMe substituted core products, and the desired product (plus recovered SM's). These side-products are not a problem as I can just add more core if necessary, which we have plenty of. The concern was about losing the alcohol, which is more valuable. So, the plan is to hit the reaction a little harder, to completion, next week.

• I intend to make use of that free tertiary alcohol compound and couple it with the core. Although shortening the linker length has previously been shown to reduce potency, the structure is now significantly different so it is worth a look and makes use of that material. (I don't want to recycle it for a second time! And we're unfortunately already a long way from being green with all the solvents used in purifying each step!)

[david1597]

The OHOH has been synthesised - we (finally) have a synthetic sample of that important potent Pfizer compound. I'm writing up the ELN with the characterisation, but for now here is a comparison of the Pfizer NMR spectrum (top) and mine (bottom). [The OCHF2 triplet appears different largely due to the 600 MHz vs 400 MHz spectrometers they were run on. OH peaks also not visible on my spectrum - they may be under the water peak]

I also have the 3,4-difluoro OHOH - we could perhaps expect this to be more potent than the original OHOH if we compare previous phenyl/difluoro pairings.

The dimethylaniline OHOH requires another re-column before I'm happy with the purity on that one.

[mattodd]

@david1597 that NMR comparison above is a thing of considerable beauty. Well done.

[mcoster]

Awesome, well done @david1597 ! What were the key factors in getting this chemistry to work?

[david1597]

Protecting groups and time. :)

It ended up a lengthy synthesis in terms of the number of steps although each individual step proceeds cleanly with easy separation. The fact that each step needs a column means it is a fairly labour intensive route.

I'm currently writing up the ELN fully and I'll add a summary of the work to the synthesis section on the wiki.

[david1597]

A brief summary on the synthesis of the OHOH's has been added to the wiki.